AppSync: Recovery of an Oracle second Gen copy fails after installing a hotfix or re-discovering the Production host.

           

   Article Number:     503634                                   Article Version: 3     Article Type:    Break Fix 
   

 


Product:

 

AppSync,AppSync 3.0,AppSync 3.0 SP1,AppSync 3.0 SP2,AppSync 3.1,AppSync 3.1 P01,AppSync 3.1 P02,AppSync 3.5

 

Issue:

 

 

After installing a hotfix or Rediscovering an Oracle Production host, Refresh of an Oracle Gen 2 Repurpose copy will fail during the mount phase. This only affects Databases on Oracle Automatic Storage Management (ASM) storage.    
   
    The Refresh will fail almost immediately with the following error:   

     
      08-22-2017 12:39:30.725 INFO [Thread-193 (HornetQ-client-global-threads-364738552)] [com.emc.archway.service.eventservice.EventServiceBean] [AppSyncserver] [] EVENT [SPP_000001]: Create 2nd gen copy phase for <DBName> beginning       
       
        08-22-2017 12:39:32.865 INFO [Thread-193 (HornetQ-client-global-threads-364738552)] [com.emc.archway.commands.application.copy.CatalogApplicationCopyCommandBean] [AppSyncserver] [] PrevMountPath: :null: :PrevDGName :null: :PrevDBName :<2nd Gen DB Name>       
       
        08-22-2017 12:39:33.334 ERROR [Thread-193 (HornetQ-client-global-threads-364738552)] [com.emc.archway.commands.serviceplan.RunReplicationCommandBean] [AppSyncserver] [] Exception Occurred: ParentResourceNotFound, Message: Parent resource of type Filesystem not found for +DATADEV/dbname/datafile/aididx.616.948802611 of type File
   
                                                             

 

 

Cause:

 

 

   

      The underlying issue is that running a gen-2 copy immediately after a host discovery causes AppSync to fail because the persistent information regarding ASM disks is blanked out by the host discovery. This only occurs for ASM as ASM disks are only discovered/mapped during a full protection that involves database discovery. Since gen-2 copies do not perform discovery and only create copies off the gen-1 copy, this error will be seen.     
     
      Re-applying the agent or installing an agent hotfix will also cause this issue as this triggers a re-discovery.    

                                                             

 

 

Change:

 

 

This can happen due to a number of situations:   

         
  1.         Running a manual rediscovery of the Oracle production host.     
  2.      
  3.         Reinstalling the Agent files on the Oracle production host.     
  4.      
  5.         Installing a hotfix on the Oracle production host.      
  6.    
                                                             

 

 

Resolution:

 

 

Workaround:    
   
    Perform a protection operation for the database in question that performs discovery/mapping operations as these operations will re-populate our persistence with the missing information for ASM disks.   
   
    One way to do this is to subscribe the database to the Bronze service plan and perform a quick protect. This will be sufficient enough to re-populate the missing persistent information. After this is complete, the newly created copy can be expired and the database can be unsubcribed from the Bronze plan.